On Fri, Nov 2, 2012 at 6:19 PM, Roman Shaposhnik <[email protected]> wrote: > Hi! > > On Tue, Oct 30, 2012 at 2:51 PM, Josh Wills <[email protected]> wrote: >> I cc'd Roman on here so we can be sure that we position the release to be >> compatible w/Bigtop. Thanks to Matthias, we have both support for .tar.gz >> release files and binary release artifacts. We're using Hadoop 1.0.3 as our >> default build target, although it is possible to do the build against the >> Hadoop 2.0.x APIs via the crunch.platform property in the top-level POM >> file. >> >> Question for everyone: is there anything else you would like to include in >> 0.4.0 that is close to completion? >> Question for Roman: is there anything else we should double-check on for >> Bigtop compatibility? > > It looks like you're in a good shape now. I will definitely track your > RC and test > it in Bigtop just to make sure. > > That said, I did notice a couple of minor things: > 1. any reason not to include examples-job.jar into assembly? The > way it stands > right now one has to jump through some hooks to make examples run.
I'll leave that one up to Matthias. > > 2. when I run AverageBytesByIP and TotalBytesByIP examples I get no > output even though the job is marked as success. Am I missing something? No, you were right. When we put fake IP addresses in the example file, the regex in the examples stopped matching. I just submitted a 2-character fix that corrects it. Thanks for finding that, and thanks for weighing in! > > Thanks, > Roman. -- Director of Data Science Cloudera Twitter: @josh_wills
